Overview
Request 913476 accepted
- Update to 2.6.0
Major changes are:
* Keras been split into a separate PIP package (keras), and its
code has been moved to the GitHub repositorykeras-team/keras.
The API endpoints for tf.keras stay unchanged, but are now
backed by the keras PIP package. The existing code in
tensorflow/python/keras is a staled copy and will be removed in
future release (2.7). Please remove any imports to
tensorflow.python.keras and replace them with public tf.keras
API instead.
* tf.train.experimental.enable_mixed_precision_graph_rewrite is
removed, as the API only works in graph mode and is not
customizable. The function is still accessible under
tf.compat.v1.mixed_precision.enable_mixed_precision_graph_rewrite,
but it is recommended to use the Keras mixed precision API
instead.
* tf.lite: Remove experimental.nn.dynamic_rnn,
experimental.nn.TfLiteRNNCell and
experimental.nn.TfLiteLSTMCell since they're no
longer supported. It's recommended to just use keras lstm
instead.
* tf.keras: The methods Model.to_yaml() and
keras.models.model_from_yaml have been replaced to raise a
RuntimeError as they can be abused to cause arbitrary code
execution. It is recommended to use JSON serialization instead
of YAML, or, a better alternative, serialize to H5.
- Major changes from 2.5.x:
* Support for Python3.9 has been added.
* The TF_CPP_MIN_VLOG_LEVEL environment variable has been renamed
to to TF_CPP_MAX_VLOG_LEVEL which correctly describes its
- Created by fusionfuture
- In state accepted
-
Package maintainers:
bnavigator and
mslacken
- Supersedes 913474
Request History
fusionfuture created request
- Update to 2.6.0
Major changes are:
* Keras been split into a separate PIP package (keras), and its
code has been moved to the GitHub repositorykeras-team/keras.
The API endpoints for tf.keras stay unchanged, but are now
backed by the keras PIP package. The existing code in
tensorflow/python/keras is a staled copy and will be removed in
future release (2.7). Please remove any imports to
tensorflow.python.keras and replace them with public tf.keras
API instead.
* tf.train.experimental.enable_mixed_precision_graph_rewrite is
removed, as the API only works in graph mode and is not
customizable. The function is still accessible under
tf.compat.v1.mixed_precision.enable_mixed_precision_graph_rewrite,
but it is recommended to use the Keras mixed precision API
instead.
* tf.lite: Remove experimental.nn.dynamic_rnn,
experimental.nn.TfLiteRNNCell and
experimental.nn.TfLiteLSTMCell since they're no
longer supported. It's recommended to just use keras lstm
instead.
* tf.keras: The methods Model.to_yaml() and
keras.models.model_from_yaml have been replaced to raise a
RuntimeError as they can be abused to cause arbitrary code
execution. It is recommended to use JSON serialization instead
of YAML, or, a better alternative, serialize to H5.
- Major changes from 2.5.x:
* Support for Python3.9 has been added.
* The TF_CPP_MIN_VLOG_LEVEL environment variable has been renamed
to to TF_CPP_MAX_VLOG_LEVEL which correctly describes its
mslacken accepted request
Thanks again, looks good.
@mslacken: review reminder
Thanks for your contribution, building tensorflow is really tough. Also I am sorry for the late reply from my side, but I was on vacations. I still see a problem, as tensorflow does not build on amd64 (I mean really Ryzen CPUs, it builds on my old Core i7), but I will try to fix this.
Local build on my machine failed, as bazel detected the local NVIDIA driver. kvm build was successful.